The present invention relates to electrochemical analysis, and more particularly to an integrated sampling system incorporating electrochemical sensors and reagents. The invention has particular utility in connection with testing lead levels in human blood and will be described in connection with such utility, although other utilities are contemplated.
It has long been possible to test both qualitatively and quantitatively for ionic materials in an aqueous sample by electrolytic means, and to record the electrical potential of deposition of the ions on an electrode. In one form of such testing known as stripping voltammetry, the ions are first deposited on an electrode and thereafter the potential is varied to strip the deposited material from the electrode and redissolve it in the sample liquid. This operation is known as stripping voltammetry and since it is ordinarily used by plating cathodically and stripping anodically to detect and measure metallic ions, it is often known as anodic stripping voltammetry. By means of anodic stripping voltammetry, it has been found possible to perform relatively quickly simple and accurate tests to measure minute traces of materials, for example, to test for the presence of informative or dangerous impurities in the human bloodstream. As a result anodic stripping voltammetry systems such as available from ESA, Inc., the assignee of the subject application, have achieved widespread field and laboratory use since at least about 1975.
The application of stripping voltammetry to testing for impurities in the human blood stream, improved electrochemical detection systems, cells, reagents and processes are the subject of several of my prior U.S. Patents, including U.S. Pat. Nos. 3,855,099, 4,201,646, 4,090,926, 4,374,041 and 4,233,031, and U.S. Reissue Pat. No. 32,920, the disclosures of which are incorporated by reference.
While anodic stripping voltammetry systems provide satisfactory results for the general population, recent studies have shown problems of individual variability affecting low blood level value (5 .mu.g %) accuracy on commercial exchange reagents/anodic stripping voltammetry systems.
Lead in a whole blood sample is distributed in a number of different compartments with a high degree of individual variable in binding constants and kinetics. Similarly, individual variability in copper levels and other endogenous oxidizable/reducible compounds can affect the envelope of the electroanalytical curve and hence the analytical value for lead.
Simplistically, generally 97-99% of lead is bound to the erythrocyte. However, within the erythrocyte there appear to be compartments both on the external surface of the cells and within the cell and sites with varying levels of binding constants. Thus, there is a level of lead typically between 2-10 .mu.g % that will not release into EDTA; a level (typically 2-8 .mu.g %) that will not release into acid; and a level (typically 0.2-2 .mu.g %) that will not release into commercial exchange reagents such as Metexchange.RTM. available from ESA, Inc. The reasons for the variability across a range of individuals is not clear, although in the specific disorder, sickle cell trait, it has been suggested that sickle cells are resistant to rupture and release and that in other cases a high concentration of Fetal Hemoglobin may imply a higher binding constant.
When lead is released from the erythrocyte to a solution matrix of whole blood residue and exchange reagents, it is bound secondarily in a number of labile ligands (sulfhydryls, phospholipids, amino acids, peptides) which in the aggregate change its apparent diffusion constant from 6.9.times.10.sup.-6 cm.sup.2 /sec to approximately 7.8.times.10.sup.-6 cm.sup.2 /sec in a 30:1 dilution to approximately 8.times.10.sup.-7 cm.sup.2 /sec in whole blood with added solid exchanger.
The analytical implications of these observations are that any technique employing direct blood measurement, such as electrochemistry or enzyme coupling, must be designed to handle the extremes of individual variability, and that it is relatively easy to be fooled into assuming technical adequacy by evaluating a technique on pools or artificially spiked samples.
The approach taken in commercial reagent/anodic stripping voltammetry to addressing the problems of lead distribution has been to dilute the blood sample, for example, by 30 times, and to provide the strongest exchange reagent possible consistent with long term sensor operation, and to operate the sensor at 1.5-1.7 half times (1 minute) to significantly reduce the lead level in solution and thus compensate for differing ligand formation constants. Essentially, the approach is a compromise among sensor longevity, sensitivity, accuracy and analysis time.
Other problems in the prior art include sample contamination, for example, due to smearing during blood drawing and sample mislabeling or mixup.